The release of potent proinflammatory mediators is not only central for mounting an efficient host response, but also bears the risk for deleterious excessive tissue-damaging inflammation. This is highlighted in severe pneumococcal pneumonia, in which the delicate balance between a robust inflammatory response to kill pneumococci and loss of organ function determines the outcome of disease. In this study, we tested the hypothesis that Krüppel-like factor (KLF)2 counterregulates pneumococci- and pattern recognition receptor-related human lung cell activation. Pneumococci induced KLF2 expression in vitro and in a murine pneumonia model. Activation of TLR2- and nucleotide-binding oligomerization domain protein 2-related signaling induced KLF2 expression in a PI3K-dependent manner. Overexpression of KLF2 downregulated pneumococci-, TLR2-, and nucleotide-binding oligomerization domain protein 2-related NF-κB–dependent gene expression and IL-8 release, whereas small interfering RNA-based silencing of KLF2 provoked an enhanced inflammatory response. KLF2-dependent downregulation of NF-κB activity is partly reversible by overexpression of the histone acetylase p300/CREB-binding protein-associated factor. In conclusion, KLF2 may act as a counterregulatory transcription factor in pneumococci- and pattern recognition receptor-related proinflammatory activation of lung cells, thereby preventing lung hyperinflammation and subsequent organ failure.
The innate immune system constitutes the first line of defense against invading pathogens. Recognition of conserved microbial molecules by pattern recognition receptor (PRR)-like transmembrane TLRs (1) or cytosolic receptors (e.g., nucleotide-binding and oligomerization domain [NOD]-like receptors) (2) generally results in the production of inflammatory mediators, subsequent recruitment and activation of leukocytes, and the initiation of the acquired immune response.
The innate immune response is characterized by the release of highly reactive mediators (3), which, in concert with products of the invaders [e.g., bacterial exotoxins, such as pneumococcal pneumolysin (4)], substantially change the microenvironment of infected tissues. As a result, significant local tissue damage and systemic distribution of potentially toxic agents and microbes may occur. Consequently, the innate immune response must be robust, rapid, and highly efficient to kill the pathogens; it also must be tightly controlled, ensuring minimal tissue damage and host survival. The vital role of this delicate balance is highlighted in pneumonia, which represents the third leading cause of death worldwide (5). On one hand, local control of infection and inflammation in pneumonia prevents the distribution of infection and inflammation within the lung and the body (pneumonia-related sepsis) (6). On the other hand, efficient oxygenation of the host must be guaranteed literally on a minute-to-minute basis, despite the ongoing combat between the host and the pathogen. However, little is known about how the innate immune response is controlled in general and in the lung in particular.
Most cases of pneumonia are due to infections with Gram-positive Streptococcus pneumoniae (7, 8). Severe pneumococcal pneumonia bears the risk for lung edema, consecutive respiratory failure, and the systemic spread of invading pneumococci (7, 8). Detection of pneumococci by transmembrane and cytosolic PRRs initiates a strong release of chemokines and cytokines (9–11), resulting in the characteristic massive infiltration of the lung by polymorphonuclear leukocytes (6, 12). Bacteria- and host cell-related products contribute to the cell and tissue injury observed in pneumococcal infection.
Krüppel-like factors (KLFs) form a subclass of ~17 zinc finger-containing DNA-binding transcription factors expressed in humans (13, 14). The different KLFs may bind with their zinc finger domains to similar DNA sequences (e.g., CACCC sequence or GT-box), and they mediate transcriptional regulation by modular activation and repression in the non–DNA-binding domains (13, 14).
KLF2 is strongly expressed in lung tissue; thus, it was first termed lung Krüppel-like factor (13, 14). Normal lung development requires KLF2, as indicated by studies using KLF2 knockout mice (15, 16). A regulatory role for KLF2 in the immune system was suggested by recent reports addressing T cell (17–19) and monocyte (18) function, as well as analysis of endothelial proinflammatory activation (20, 21). Moreover, Pseudomonas aeruginosa exoenzymes S and Y were recently suggested to induce KLF2 and KLF6 expression in airway epithelial cells (22), and KLFs were considered targets of bacterial exotoxins (23).
Therefore, we tested the hypothesis that KLF2 expression contributes to the control of proinflammatory gene expression in pneumococcal infection.
Materials and Methods
SB202190, SP600125, U0126, wortmannin, PMA, and Ly294002 were purchased from Calbiochem (Merck, Darmstadt, Germany). Muramyldipeptide (MDP) and macrophage-activating lipopeptide-2 (MALP-2) were from Alexis Biochemicals (Lörrach, Germany), and NF-κB essential modulator-binding domain peptide (IKK-NBD) was from BIOMOL (Plymouth Meeting, PA). All other chemicals used were of analytical grade and were obtained from commercial sources.
Pneumonia in mice was induced by the encapsulated serotype 3 strain S. pneumoniae PN36 (National Collection of Type Cultures 7978), as described (24, 25). S. pneumoniae R6x (and R6xΔply; pneumolysin deficient) used for in vitro studies are the unencapsulated derivatives of serotype 2 strain D39 (10, 26). Pneumococci were cultured as described previously (25, 26).
Human bronchial epithelial cells (BEAS-2B) (C. Harris, National Institutes of Health, Bethesda, MD) were cultured in keratinocyte serum-free medium supplemented with retinol acid, epidermal growth factor, bovine pituitary extract, and epinephrine. HEK293 cells were cultured as described (9). Primary small airway epithelial cells were cultured in small airway cell basal medium (both from Cambrex, Taufkirchen, Germany). Cells were infected with pneumococci using the indicated CFU and incubated at 37°C and 5% CO2.
Mouse pneumonia model
For analysis of KLF2 induction in vivo, pneumococci-infected mice lungs were snap-frozen in liquid nitrogen and pulverized. Cell or lung homogenates were lysed, subjected to SDS-PAGE, and blotted on Hybond-ECL membrane (GE Healthcare, Munich, Germany). Membranes were exposed to KLF2 Ab (Santa Cruz Biotechnology, Heidelberg, Germany, or L. Glimcher, Harvard School of Public Health, Boston, MA), ERK2, actin, cyclooxygenase (COX)2 (Santa Cruz Biotechnology), focal adhesion kinase (Millipore, Schwalbach, Germany), TLR2 (Acris Antibodies, Hiddenhausen, Germany), NOD2 (ProSci, Poway, CA), and pPI3K (Cell Signaling Technology, Frankfurt, Germany); subsequently incubated with secondary Abs (IRDye 800- or Cy5.5-labeled, Rockland, Gilbertsville, PA); and detected using an Odyssey infrared imaging system (LI-COR, Bad Homburg, Germany) (9, 26).
Confluent BEAS-2B cells were stimulated for 15 h. Supernatants were collected and processed for IL-8 ELISA, according to the manufacturer’s instructions (BD Biosciences, Heidelberg, Germany).
Subconfluent HEK293 cells were transfected using the calcium phosphate precipitation method (BD Clontech, Palo Alto, CA), and BEAS-2B cells were transfected using Fugene 6 (Roche Applied Science, Mannheim, Germany). Luciferase-dependent reporter plasmids [NF-κBluc (9, 26), IL-8luc (M. Kracht, Rudolf-Buchheim-Institut für Pharmakologie, JLU Gießen, Germany), KLF2luc (S. Choksi, National Institutes of Health), pRL-TK (Renilla Luciferase Control Reporter Vector) (Promega, Mannheim, Germany), p300/CREB-binding protein-associated factor (PCAF) (Y. Nakatani, National Institute of Child Health and Human Development, Bethesda, MD), human TLR2 (Tularik, San Francisco, CA), human NOD2 (9), and KLF2 and pcDNA (both from S. Banerjee, Harvard Medical School, Boston, MA)], were used as indicated. Luciferase activity was measured using a Luciferase Reporter-Gene Assay and normalized for transfection efficiency with values obtained with the Renilla Luciferase Assay System (both from Promega).
RNA interference in BEAS-2B cells
Control nonsilencing small interfering RNA (siRNA) (sense 5′-UUCUCCGAACGUGUCACGUtt-3′, antisense 5′-ACGUGACACGUUCGGAGGAGAAtt-3′), TLR2 (sense 5′-GACUUAUCCUAUAAUUACUtt-3′, antisense 5′-AGUAAUUAUAGGAUAAGUCta-3′), NOD2 (sense 5′-GGAAUUACCAGUCCCAUUGtt-3′, antisense 5′-CAAUGGGACUGGUAAUUCCtg-3′), and KLF2 siRNA (sense 5′-CGAUCCUCCUUGACGAGUUtt-3′, antisense 5′-AACUCGUCAAGGAGGAUCGtg-3′) were from Ambion (Huntingdon, Cambridge, U.K.). siRNA-lipoplexes were generated with AtuFECT01 for transfection (Silence Therapeutics, Berlin, Germany), using 2 μg siRNA per 106 cells (27).
Plasmid-transfected BEAS-2B cells were stimulated for 1 h before being processed for chromatin immunoprecipitation analysis, as described previously (26). Immunoprecipitations were carried out with p65 and KLF2 Abs (Santa Cruz Biotechnology). The following primers for il8 were used: sense, 5′-AAGAAAACTTTCGTCATACTCCG-3′; antisense, 5′-TGGCTTTTTATATCATCACCCTAC-3′.
Data are shown as the mean ± SEM for at least three independent experiments. One-way ANOVA was used, and the main effects were compared using the Newman–Keuls posttest.
Pneumococci induce KLF2 expression in vitro and in vivo
First, we analyzed the induction of KLF2 by S. pneumoniae. Pneumococci induced time- (Fig. 1A) and dose- (Fig. 1B) dependent expression of KLF2 in BEAS-2B cells, as well as in primary small airway epithelial cells (Supplemental Fig. 1). Enhanced activation of a KLF2-luciferase reporter was measured in BEAS-2B cells infected with R6x and pneumolysin-deficient R6x (Δply) (Fig. 1C). Further, we noted increased KLF2 expression in whole-lung tissue lysates from mice with pneumococcal pneumonia (Fig. 1D).
KLF2 expression by pneumococci requires TLR2 and NOD2
Recognition of pneumococci by TLR2 and NOD2 was shown to contribute to the proinflammatory activation of host epithelial cells by pneumococci (9–11). Accordingly, the TLR2 ligand MALP-2 (Fig. 2A), as well as activation of cytosolic NOD2 by MDP (Fig. 2B), increased KLF2 expression by BEAS-2B cells. The pneumococcal-derived TLR4 ligand pneumolysin induced KLF2 expression only at the highest (1 μg/ml) dose. Of note, the pneumolysin-negative mutant R6xΔply was as effective as the pneumolysin-containing R6x strain with respect to KLF2 induction (Fig. 2C), indicating that pneumolysin and TLR4 play no major role in pneumococci-related KLF2 expression. We used validated TLR2 and NOD2 siRNA to confirm the roles of TLR2 and NOD2 in KLF2 expression (Supplemental Fig. 2). Silencing of TLR2 and NOD2 expression reduced pneumococci-related KLF2 induction (Fig. 2D).
We verified these results by using HEK293 cells transiently transfected with control vector (Fig. 3A) or human TLR2-expressing (Fig. 3B) or human NOD2-expressing (Fig. 3C) plasmids. Heat-inactivated pneumococci induced the activation of a KLF2-dependent reporter gene in TLR2-overexpressing (Fig. 3B) and NOD2-overexpressing (Fig. 3C) cells and stimulated the activation of an NF-κB–dependent reporter construct (Supplemental Fig. 3). Comparable results with regard to KLF2 and NF-κB reporter gene activation were obtained by using MDP for stimulation of NOD2 or MALP-2 for TLR2 activation. Overall, transmembrane TLR2 and cytosolic NOD2 induced KLF2 expression in these models. However, in mock-transfected cells, a slight heat-inactivated pneumococci- and MDP-dependent activation of NF-κBluc (Supplemental Fig. 3B), but not KLF2luc (Fig. 3A), was observed.
PI3K, but not p38 MAPK, ERK, or JNK, regulates KLF2 expression in lung cells
Mouse fibroblasts deficient in p38 MAPK did not show TRAF2-dependent KLF2 expression (28). In addition, MAPKs (p38 MAPK, ERK, and JNK) and NF-κB are known to be important in PRR signal transduction, in general (1, 2), and in pneumococci-related activation of lung cells, in particular (10, 12, 26). Therefore, we assessed the role of these signaling molecules in pneumococci-related KLF2 expression. Although chemical inhibitors (SB202190 for p38 MAPK, U0126 for MEK1, and SP600125 for JNK) and the NF-κB inhibitor IKK-NBD reduced the pneumococci-related expression of COX2 (Fig. 4A, 4B), neither the MAPK inhibitors (Fig. 4A) nor IKK-NBD (Fig. 4B) altered KLF2 protein expression in lung epithelial cells.
PI3K was shown to regulate fluid shear stress-related KLF2 expression in endothelial cells (21). When using two chemically unrelated inhibitors of PI3K, wortmannin and Ly294002 (Fig. 4C), we noted that both inhibitors reduced pneumococci-related KLF2 expression in BEAS-2B cells. Also, Ly294002 reduced the activation of a KLF2 reporter gene in pneumococci-infected lung epithelial cells (Fig. 4D). Furthermore, activation of TLR2 (MALP-2) and NOD2 (MDP), as well as stimulation of cells with pneumococci, induced the phosphorylation of regulatory Tyr199 of PI3K subunit p55 in lung epithelial cells (Fig. 4E–G). However, a constitutive phosphorylation of p85-Tyr458 was noted in these cells (data not shown).
KLF2 suppressed pneumococci-related NF-κB activation and IL-8 expression
NF-κB–related gene expression is a central event in pneumococci- and PRR-induced activation of the host innate immune response (9, 10, 26, 29). To test whether KLF2 expression contributed to the modulation of pneumococci-related NF-κB gene transcription, we analyzed the effect of enhanced KLF2 expression on pneumococci-induced activation of an NF-κB (Fig. 5A) and IL-8 (Fig. 5B) reporter construct in transfected BEAS-2B cells. KLF2 overexpression significantly reduced pneumococci- and PMA-related activation of both reporter genes (Fig. 5A, 5B). In contrast, knockdown of KLF2 expression by KLF2 siRNA (Supplemental Fig. 4A) augmented the pneumococci-related activation of the NF-κB reporter in TLR2- or NOD2-overexpressing HEK293 cells (Fig. 5C), as well as in R6x-stimulated BEAS-2B cells (Supplemental Fig. 4B). Furthermore, siRNA-mediated suppression of KLF2 significantly increased IL-8 release by BEAS-2B cells compared with cells exposed to pneumococci alone (Fig. 5D).
KLF2-related NF-κB repression due to competition of histone acetylases
Prior studies in LPS-stimulated monocytes and in IL-1β–exposed endothelial cells indicated that KLF2 did not directly interfere with the canonical NF-κB pathway (18, 30). In contrast, Lin et al. (31) reported a KLF2-related reduction in NF-κB DNA-binding by endothelial cells activated with thrombin. To determine whether KLF2 influenced the canonical NF-κB pathway in our study, we investigated the recruitment of the NF-κB p65 subunit to the il8 promoter in BEAS-2B cells overexpressing KLF2 1 h postinfection with R6x; KLF2 had no effect on the recruitment of the NF-κB p65 subunit to the il8 promoter (Fig. 6A). In endothelial cells, histone acetylases were recruited from NF-κB to KLF2, thereby terminating NF-κB activation and proinflammatory gene expression (18, 30). KLF2-dependent downregulation of MALP-2–related (Fig. 6B) or MDP-related (Fig. 6C) NF-κB reporter gene activation is reversed by cotransfection of the histone acetylase PCAF by the NOD2 ligand but not by the TLR2 ligand.
In this study, we examined the role of KLF2 in the innate immune response of lung epithelial cells infected with S. pneumoniae. Activation of epithelial cells by pneumococci, TLR2, or NOD2 ligands induced the expression of KLF2 in a PI3K-dependent manner. Gain- and loss-of-function experiments showed that KLF2 expression dependent on TLR2 stimulation, as well as NOD2 stimulation, reduced NF-κB activation and subsequent NF-κB–related IL-8 expression. One possible underlying mechanism seems to be the competition between KLF2 and NF-κB for histone acetylases. Overexpression of the histone acetylase PCAF reversed the KLF2-dependent reduction of NF-κB activity in MDP-stimulated cells. However, PCAF overexpression in MALP-2–stimulated cells is not sufficient to recover NF-κB activity (see Fig. 7). Taken together, these results suggest that KLF2 acts as an important counterregulatory transcription factor in the proinflammatory response during pneumococcal lung infections, thereby preventing deleterious lung hyperinflammation and subsequent organ dysfunction.
In infections, initiation of an efficient innate immune response and, thus, elimination of microbial invaders, is vital for the host. In contrast, overwhelming inflammation may lead to tissue damage and deleterious organ dysfunction. Accumulating evidence indicates that KLF2 may act as a modulator of endothelial- (20, 21, 30), monocyte- (18), and T cell- (17, 19) related immune responses. By using pneumococci-mediated lung cell activation as a model, we tested the hypothesis that KLF2 expression contributes to PRR-related regulation of the innate immune response.
Exposure of lung epithelial cells to Gram-positive pneumococci resulted in strong protein expression of KLF2 under in vitro and in vivo conditions, as shown in infected mice lungs. The delayed expression in vivo compared with in vitro is a common observation in our studies (26). This may due to different infection doses used in vitro and in vivo, adhesion of bacteria in the upper respiratory tract, or inhomogeneous distribution of bacteria in the lung. Thus, the measurement of the KLF2 signal in Western blot may not reflect earlier KLF2 expression in severely infected areas of the lung. Furthermore, immune cells, such as monocytes and macrophages, are known to express KLF2, and these cells were recruited to the lung in pneumonia. Therefore, a detailed immunohistological analysis of KLF2 expression in infected human and mice lung tissue in additional studies is needed to clarify the cell-specific expression pattern of KLF2 in the infected lung. The Gram-negative lung pathogen P. aeruginosa also induced KLF2 expression in airway epithelial cells (22), suggesting that KLF2 induction is not restricted to pneumococci or Gram-positive bacteria. Moreover, global gene-expression analysis showed increased KLF2 mRNA expression in Yersinia enterocolitica-infected cells (32, 33). These data strengthen the hypothesis that KLF2 is involved in the regulation of the host response to bacterial infections.
To gain more insight into KLF2 induction, we assessed whether recognition of pneumococci may contribute to KLF2 expression. In epithelial cells, pneumococci were recognized by transmembrane TLRs (e.g., TLR2 and TLR4) (6, 10–12), as well as by cytosolic NOD2 (9, 12). We found increased KLF2 expression in cells exposed to TLR2 and NOD2 ligands, as well as in cells exposed to very high concentrations of the pneumococcal-derived TLR4 ligand pneumolysin. However, we showed in a previous study that R6x failed to produce a sufficient amount of pneumolysin to activate TLR4-dependent NF-κB activity in epithelial cells (10), indicating that the TLR4 pathway has no impact on our experimental setting with respect to KLF2 induction. Nevertheless, we cannot exclude that other unknown PRRs may contribute to pneumococci-induced KLF2 expression. NOD2 was previously reported to initiate proinflammatory cell activation and to have counterregulatory properties. For example, Rac1- and β-PIX–related negative regulation of NOD2 resulted in a reduction in IL-8 release in MDP-stimulated monocytes (34). Furthermore, ligand binding to NOD2 may downregulate TLR-related cell responses (35). In addition, to the best of the authors’ knowledge, this is the first study to report a direct correlation between TLR2 activation and downregulation of NF-κB activity.
At least in lung cells, forced KLF2 expression seems to be a common phenomenon in PRR-related cell activation. Because exposure of endothelial cells to proinflammatory IL-1β resulted in decreased KLF2 expression (30), a signal- and cell-specific response has to be considered with regard to KLF2 expression.
Several kinases are important for the regulation of pneumococci (10, 26, 36), in particular, as well as for PRR-related expression of proinflammatory genes, in general (1, 2, 12). In our experiments, we noted that pneumococci, as well as stimulation of TLR2 and NOD2, induced the phosphorylation of the regulatory p55α subunit of PI3K, whereas the p85α subunit showed a constitutive high phosphorylation, indicating enzyme activation. The p55α and p85α subunits of PI3K were reported to play an essential role in the development and function of the immune system in vivo (37, 38). Blocking of PI3K by two chemically unrelated inhibitors reduced the pneumococci-mediated expression of KLF2. By analyzing shear stress-induced KLF2 expression in endothelial cells, Huddleson et al. (21) noticed PI3K-dependent acetylation of histones H3 and H4 at the klf2 promoter, which paves the way for chromatin remodeling and subsequent klf2 gene expression. Together, these results suggested a critical role for PI3K in the regulation of KLF2 expression. PI3K-related activation of KLF2 expression may, at least in part, explain the anti-inflammatory effects of PI3K (39). In contrast, although MAPKs p38, ERK, and JNK participated in pneumococci-induced expression of various inflammatory regulators in lung epithelial cells (9, 26, 40), and p38 MAPK abolished TNFR-associated factor 2-dependent KLF2 expression in fibroblasts (28), these kinases did not interfere with pneumococci-mediated KLF2 protein expression in epithelial cells.
Stimulation of NF-κB–related gene expression is an integral issue in PRR-related cell activation and essentially contributes to the orchestration of the innate immune response (1, 2). Liberation of the NF-κB–dependent CXC chemokine IL-8 is of particular importance for the recruitment of polymorphonuclear leukocytes into the airspace in pneumonia (41, 42) and is liberated by pneumococci-infected lung epithelial cells (10, 26). Although forced expression of KLF2 reduced expression of an NF-κB and IL-8 reporter, siRNA-mediated KLF2 knockdown increased NF-κB reporter activity, as well as IL-8 liberation, in pneumococci-infected cells. Negative regulation of NF-κB activation (30, 31) and subsequent IL-8 expression (18, 43) by KLF2 was described previously, but the mechanism has not been completely elucidated.
As we reported previously, pneumococci-dependent stimulation of TLR2 (10, 26) and NOD2 (9) lead to activation of the canonical NF-κB pathway. The interference of KLF2 with the NF-κB pathway is controversial, and it seems to vary with different stimuli. Although, for example, KLF2 did not influence NF-κB DNA binding after stimulation of endothelial cells with IL-1β (18), the same group showed that after stimulation with thrombin, NF-κB DNA binding was strongly reduced by KLF2 in the same cell type (31). In our study, KLF2 overexpression had no impact on the recruitment of the NF-κB subunit p65 to the il8 promoter. Some investigators hypothesized that the termination of NF-κB–dependent gene transcription (e.g., IL-8) depends on the competition between NF-κB and KLF2 for histone acetylases, such as PCAF (18, 30, 44). In an NF-κB reporter assay, we could overcome the KLF2-dependent NF-κB inhibition via overexpression of PCAF in MDP-stimulated, but not in MALP-2 stimulated, BEAS-2B cells. These results suggested that histone acetylases may play an important role in NOD2-induced, but not TLR2-induced, KLF2-dependent abolishment of NF-κB activity. The underlying mechanism by which KLF2 interferes with TLR2-dependent NF-κB signaling remains unclear and needs further investigation. Because acute lung hyperinflammation may lead to severe organ dysfunction, increased KLF2 expression may act as a counterregulatory signal, balancing the acute inflammatory response in pneumonia.
The lung alveolus is essentially formed by lung epithelial and endothelial cells. Importantly, KLF2 is suggested to act as a potent negative regulator of inflammatory signals in activated endothelium (20, 21, 30). Thereby, KLF2 may dampen the release of proinflammatory mediators in lung epithelium and endothelium; furthermore, it may act as a barrier-stabilizing molecule, as shown for vascular endothelial growth factor-mediated edema formation (45). It would be of interest to analyze how KLF2 expression contributes to the regulation of edema formation induced by pathogen-released agents, such as pneumolysin (24, 46), as well as potent endogenous mediators, such as TNF-α, thrombin, or oxygen radicals, in pneumonia (42, 47–50). KLF2 knockout analysis in mice is hampered by early lethality between E12.5 and E14.4 through defective regulation of cardiac output and heart failure (16, 51), and KLF2+/− mice die after birth, showing abnormalities in their lung development (15). The generation of inducible knockout models (e.g., cre-lox) would be very helpful to investigate the impact of KLF2 in pneumococcal pneumonia in vivo. Furthermore, organ-specific cre or adoptive cell transfer/chimeric mice are needed to dissect the role of lung epithelial- versus immune cell-related KLF2 in pneumonia. Although in vitro experiments and mice models contributed significantly to the understanding of pneumococcal pneumonia, the disease is naturally a human-specific one; thus, additional studies using infected human lung tissue are necessary to verify the role of KLF2 in humans. On the molecular level, in-depth analysis of KLF2 induction by whole pathogens and the role of the various signaling pathways induced by the transmembrane (1, 29) and cytosolic receptors (2) in different cells and tissues is needed. Further functional analysis of KLF2 interaction with these pathways will help us to understand how KLF2 contributes to gene regulation in innate immunity. For example, in addition to regulating NF-κB–dependent gene transcription (18, 30), the role of KLF2 in IFN-inducing signaling pathways (52) is unknown. In particular, the interaction of KLF2-related responses with other counterregulatory molecules of NF-κB–related gene expression in acute inflammatory processes requires further investigation (53, 54).
It is noteworthy that recent reports also indicated an important regulatory role for KLF2 in the adaptive immune response. For example, essential steps in T cell activation and migration seem to be dependent on KLF2 function, thereby linking this transcription factor to the adaptive immune system (17, 19).
In conclusion, this study indicated that KLF2 may act as a counterregulatory transcription factor in pneumococcal pneumonia. Activation of TLR2 and NOD2 resulted in PI3K-dependent KLF2 expression in lung cells, dampening NF-κB–related gene expression. Therefore, KLF2 expression may be vital to prevent deleterious hyperinflammation in pneumonia.
We thank J. Hellwig, D. Stoll, A. Kühn, and F. Schreiber for excellent technical assistance. Parts of this work will be included in the M.D. thesis of R.S.
Disclosures The authors have no financial conflicts of interest.
This work was supported by grants from Bundesministerium für Bildung und Forschung-BMBF-FORSYS-Partner (to B.S.); BMBF-Network PROGRESS, DFG HI-789/6-1, EU-Network CAREPNEUMO (to S. Hippensteil); BMBF-Network PROGRESS (to N.S.); EU-Network CAREPNEUMO (to S. Hammerschmidt); OP-87/1 (to B.O. and M.W.); and the Deutsche Gesellschaft für Pneumologie (to J.Z., S. Hippensteil, and B.O.).
The online version of this article contains supplemental material.
Abbreviations used in this paper:
control small interfering RNA
NF-κB essential modulator-binding domain peptide
nucleotide-binding oligomerization domain
p300/CREB-binding protein-associated factor
pattern recognition receptor
relative luciferase unit
small interfering RNA